AbstractObjective:To examine the association between economic residential segregation and food environment.Design:Ecological: Food stores categorised according to the NOVA classification were geocoded, and absolute availability was calculated for each neighbourhood. Segregation was measured using local G_i^* statistic, a measure of the sd between the economic composition of a neighbourhood (the proportion of heads of households in neighbourhoods earn monthly income of 0 to 3 minimum wages) and larger metropolitan area, weighted by the economic composition of surrounding neighbourhoods. Segregation was categorised as high (most segregated), medium (integrated) and low (less segregated or integrated). A proportional odds models were used to model the association between segregation and food environment.Setting:Belo Horizonte, Brazil.Participants:Food stores.Results:After adjustment for covariates, neighbourhoods characterised by high economic segregation had fewer food stores overall compared with neighbourhoods characterised by low segregation (OR = 0·56; 95 % CI (0·45, 0·69)). In addition, high segregated neighbourhoods were 49 % (OR = 0·51; 95 % CI (0·42, 0·61)) and 45 % (OR = 0·55; 95 % CI (0·45, 0·67)) less likely to have a high number of food stores that predominantly marketed ultra-processed foods and mixed food stores, respectively, as compared with their counterparts.Conclusions:Economic segregation is associated with differences in the distribution of food stores. Both low and high segregation territories should be prioritised by public policies to ensure healthy and adequate nutrition as a right for all communities. The former must continue to be protected from access to unhealthy commercial food outlets, while the latter must be the locus of actions that limit the availability of unhealthy commercial food store.